Everyday AI Podcast – An AI and ChatGPT Podcast - EP 11: AI and Digital Transformation: What's Next?

Episode Date: May 9, 2023

Aimann Rasheed joins Everyday AI to discuss what's next in the digital transformation space. What role will AI play? Time Stamps:[00:00:50] IBM's Watson X: From Jeopardy to Enterprise and Beyond[00:04:00] The Debate on the Widespread Use of AI[00:05:20] AI Ordering Systems: The Future of Fast Food?[00:07:44] Google's Caution with AI[00:09:30] Middle-sized Businesses Prioritizing AI for Efficiency[00:12:02] Unleashing Innovative AI Solutions: The Key Challenges[00:14:07] The Future of AI Training: Using APIs[00:16:00] The Future of White-Collar Jobs with AIFor full show notes, head to YourEverydayAI.comSend Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info)

This is the Everyday AI Show, the Everyday Podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live in Adobe Firefly, the All In One Creative AI Studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. Will we even realize what work will look like in a few years?

I'm not sure, but that's one of the things that we're going to be talking about today in everyday AI, your daily podcast, live stream, and newsletter going over the latest AI news. So today we're going to talk about a lot of things, really excited for the show. We're going to talk about IBM hasn't given up on Watson. Leadership qualities are changing with AI. And is an AI chatbot going to be coming to a drive-through near you? So as you'll see, if you're joining us live, we have Aymann, Rashid. He is the technology leader of digital transformation at Eisner, Amber.

Amen, thank you so much for joining us. Good to be here, Jordan. All right. So let's start on IBM and Watson. So as most, well, maybe people don't know. So, you know, IBM was really early to the party with their original AI supercomputer called Watson. They kind of sold that piece. So they're bringing back Watson X, which is just their second attempt to go all in on AI and, you know, helping to provide that to

clients. I think the thing that's really interesting here is they're partnering with Hugging Face. So for the everyday person, you know, if you've never heard of Hugging Face, they are essentially an open source. I don't know if you'd say alternative, but they are an open source alternative to OpenAI and chat GPT, and they have a whole suite of programs. But, you know, Avan, I know you have some experience actually working with, you know, IBM Watson. So what are your thoughts on this kind of reinvigoration of Watson X? Yeah, so I don't know a whole lot about Watson X just yet because, you know, they recently announced. But, you know, I don't know if you remember, maybe the audience remembers almost 10 years ago Watson debuted on Jeopardy.

Do you remember that episode? Yeah, I remember. Yep. And back then, it was very like pie in the sky out of reach. This is like bleeding edge technology. No one's going to, it's never going to materialize anytime soon. But actually, believe it or not, Watson and Watson's suite of solutions has been around on the enterprise level for a long time. And, you know, this is something that I didn't realize. But when you go on, for example, I think it was American Airlines and you talk to their chat agent, it is an AI. And it's embedded so seamlessly that you can't even tell, right? So I think a lot of these enterprises have been using IBM's products on the enterprise level

because they offer like a hybrid, which means like you can host it on the cloud or on-premise servers. And they have what's called a cloud pack. And it's kind of like an all integrated foundational service with Watson assistant on top of it. You can train the model. You can use it to consume like a database of information. and then people can chat with the assistant to get different information about your website or get help with whatever. So IBM has been around for a while.

I think Watson X is meant to be like more consumer facing. So maybe it's meant to be a little bit more user friendly, a little bit more within reach. It's less enterprise grade. So hopefully that means for smaller businesses, it's a little bit easier to, although IBM enterprise solutions aren't necessarily expensive. Right. But it's just a different market entirely. So yeah, I need to definitely read up more on it.

It's exciting. And they've been doing great work and not getting a whole lot of credit for it. And I think that's, I think everyone except Open AI has been kind of left behind from a marketing perspective, but not necessarily from a capability perspective. And that's why we're super early in the hype cycle. For sure. It should be interesting to see. So yeah, second second kind of big story for for today on Everyday AI is a new study that was written about in Forbes, but the study is from Citrix. Just talking about leadership qualities in general are changing with the future of AI. So I found this interesting. The results of this survey, they said in 10 years, 86% of people feel that AI will be widespread, kind of in their day-to-day work. I was actually thrown off by that number. Avid, what are your first takes on just hearing that number in 10 years, 86% feel that way? I mean, that's pretty high.

86%. I mean, that's 14% away from 100%. I guess the question is, what does widespread mean, right? I mean, everyone knows that AI is already out there. The question is, how widespread is it going to be? That remains to be seen, I think, There is on both sides of thoughts, you know, I would say maybe the 14% that said, no, don't really understand AI all that well. Right.

Or maybe they think it's going to have limited uses. But people are, I mean, the younger generation is using it a lot right now. And over time, that's just going to get closer to 100%. But I think it's partially ignorance. Yeah. Yeah. And maybe one thing that will bring it to 100%. Kind of our third news story of the day.

So Wendy's is working with Google to essentially. integrate current language models into their offerings in their drive-through. So I know these have been around for a while in different phases of testing, but I think this might be the first time that at least I've seen publicly a company like Google working with a fast food chain, you know, just training it on their terminology. The example I end the story today is, you know, saying, hey, JBC is Junior Breaking Cheeseburger, right? What are your thoughts on maybe an improved version of this AI coming to a drive-through. I could have sworn I heard that McDonald's or another fast food chain has an AI ordering system.

Yeah, I think it was McDonald's. And they obviously, they only piloted it in like one or two locations. But it was pretty bad. But, you know, that could be for several reasons. But I think that it's not all that far-fetched to believe that it's going to become the norm for ordering systems and to free up the humans so that they can cook instead of taking down orders. I think it's going to happen. When it does happen, it's going to be ubiquitous, and it's going to work well enough where no one's going to even notice. Yeah, it'll be sad. That piece on my resume

working the drive-thru at Culver's will no longer get me much clout. So, all right, so let's transition a little bit. So, you know, Google has their big I-O conference tomorrow, I believe, and they're supposed to be announcing just a lot of updates to their current, to their current AI offerings and their suite of products. Personally, I feel Microsoft has really been pushing the envelope. But, you know, even in your role in digital transformation, you know, what do you think that, you know, I know we don't know exactly what Google will announce, but, you know, how do you feel with just the big announcements,

from companies like Google and Microsoft and AMD and Invidia. It seems like all the biggest companies are saying publicly on their earnings call everywhere, they're going all in. It seems on AI. So, you know, what does that mean both for everyday people like you and me, but also even, you know, working in digital transformation? What does this, you know, what does this mean when all these companies go all in? Yeah. So word on the street is that Google has a lot more than it's showing us and they're just taking precautions. And I also heard that, you know, leadership, you know, put their foot down and was like, you know, we can't be so risk averse. I think there's a lot of trepidation because of, you know, potential lawsuits based off the information AI gives you.

It's a very great area because AI is a loose canon, right? I mean, you train it on data set and it says whatever it thinks it's right. And that's, I mean, that kind of represents the good and the bad of human nature, right? We're all loose cannons. We're all creative minds and we all have a lot we can say that can either be wrong or offensive or whatever. Now package that in a commercial setting and let it loose and you have no idea what's going to happen. So I think Google is being very cautious, but that could work against them if they take too long. So I think it'll be interesting. I think it's obviously good for the consumer that more competition comes out because I'm sure their model is different than chat GPT or GPT4 rather,

the GPT4 model in terms of the types of responses you get and maybe what it's good at versus what it's bad at. But I hope they really come out of the gates swinging and not hold back. Because that would be best for everybody. Yeah, and you bring up an interesting point because I do believe that you can make the argument, And obviously the technology is there, but there's obviously things holding them back.

You know, even, you know, yesterday on the show that we talked about, you know, Apple CEOs, you know, specifically even saying that, you know, taking a specific, like a dedicated and slow approach. Do you think that, you know, even in your role at Eisner-Ampard, do you think that, you know, larger enterprise companies are taking a similar approach? Are they also taking a like, hey, we know the technology's there, but let's wait? I wouldn't say they're saying let's wait. I think they're exploring conceptually and not implementing just yet. I would say middle, so we work with middle size and enterprise size businesses,

and I would say the middle ones are preoccupied with improving their operational efficiency and trying to automate certain things that are slowing them down and getting on the right systems. So especially if they're pivoting towards enterprise level in terms of their annual revenue. enterprise companies have a lot more at stake because, you know, there's already a company, even IBM itself is replacing a decent amount of its workforce with AI. So if you're not doing it on an enterprise level, you're going to quickly become obsolete. So they do have to pay attention. The question is, to what extent can they actually implement AI and either increase the productivity on a per worker basis or reduce. reduce the amount of labor they need or the amount of resources they need.

And nobody likes to talk about that. But that is disinflationary. It would help with the economy overall if more companies did it. But again, it is, it's new. And these tools are still kind of coming out and being refined. And it's going to be the trailblazers that really set the stage and show a good example for everybody else. Yeah. Yeah.

Yeah. I mean, you mentioned something, you know, that we talked about, I think last week on the show. So, yeah, IBM saying, I think 7,800. under jobs that they were not going to fill them and instead were going to be essentially filled by AI. You know, and then you also talked about with, with, you know, these mid to enterprise size clients, like what is the extent that they either are or should be implementing AI? So, so not asking you to, you know, look into the future, but if you had to look into

the crystal ball today, what, what could it look like? I mean, can medium size and enterprise clients like, like the IBMs of the world and And those smaller, obviously, you know, let's say, you know, Fortune 500s and, you know, companies doing, you know, 20 million, 50 million, 100 million a year. So you're smaller and medium-sized clients. Can they take moves like IBM and say, hey, we can, this 20% of our work, this could go to AI? They can. Now, the challenge here is that they have to get really creative, dig deep, and understand

what specific use cases or what business case can they make for. implementing AI because the problem, I guess, with AI is that it's not a cookie cut of solution. You can't just take AI and toss it on your desk and say, hey, we've implemented AI. It's very much unique to what they need. And they have to do some engineering and some creative thinking and design thinking. We conduct enterprise design thinking workshops where we take the client through a journey that helps them understand what they need to implement. from a modernization perspective or a digital transformation perspective, that's very important so that they can understand, you know,

what the ROI is from doing this, you know, uplift. With AI, it's a lot more, there's a lot more creativity required because AI is not quite there in terms of general intelligence. I mean, it's not even close to general intelligence, in my opinion, but it can fool you into believing that. So you have to definitely create use cases. that are well suited for the current models of AI or invest a lot of money into a totally separate, totally different, you know, supervised or unsupervised learning using neural networks that it's

going to be very resource intensive. So there's, if you want something out of the box, you have to know exactly what you're doing. If you want something that's unique, you need to know what you're doing and you need a lot of money. So those are the two bottlenecks, I would say. Which, you know, of those two options, so kind of the, the cookie cutter out of the box versus the companies, you know, you've kind of started to mention, you know, some machine learning and, you know, supervised, unsupervised. Which of those two routes do you think is going to be more widely traveled in the near future? You know, people just saying like, hey, we're going to go to the cookie cutter approach to this AI or, hey, we're going to invest and we're going to build

something out specifically for us. Neith. I think that the, most common approach will be to just use large language models and use an existing API to prompt it. I would say that's a class called generative AI, which is just helping AI fill in the blanks, as opposed to teaching AI something brand new and then asking it to literally take over a certain part of your company with like auto GPT and some of the AI agents, we have kind of different models where we can continuously prompt. And that's, again, that's large language models. When you get into like the supervised and unsupervised learning and building your own deep learning

system, you have to have very, very specific game plan. It's a lot harder and it's a lot more out of reach for now until the hardware catches up and becomes extremely cheap. So I think that AI training as a service will eventually become a thing. And eventually it'll be such that, you know, you use an API to feed a data. For example, Amazon you can use like their streaming platform. I think it's called kinetic to basically firehose data into an AI training model that's simple.

But right now, it does still require a level of expertise that not a lot of companies have. But if they don't get on that, especially on the enterprise level, they will be left behind. So it's do or die. This is kind of the inflection point for the industries. Yeah. I do think that is kind of where the people

at least who understand AI and who are following it, they are starting to say, hey, this is a do or die time for big companies. So I said I wasn't going to make you looking at the crystal ball, but now I'm just going to put you on the spot here. So as we are,

as we are low on time. Give me your quick hot take on AI. Are we all going to be out of work? Are we all going to be just working on prompt engineering? You know, the average everyday person who just has a dust job, what is their role going to look like in five years? It's funny you say that. So there's a, okay, so I'll give you the pessimistic and the optimistic view, okay?

Love it. Pessimistic view is most white collar jobs. where you're shuffling data around is going to be more, it's going to be more a scenario where a company will say, like, oh, do we need a human for this AI to help manage it or not? Otherwise, the AI can just kind of function on its own. You have kind of an orchestrator at the top, like a C-level or maybe you still need some middle management, but not nearly as much management as you need now.

And then that AI will take on the role of checking in with individuals, individual resources. Imagine you get a video call at Jordan right now, and it's from, I don't know, Brad Pitt. Right? And it's a fake Brad Pitt, but it's so good. It's a deep fake. His voice is real, and he's talking to you about what you did. He's basically your manager, and he's also your therapist, by the way, because he can do whatever.

And he's just making sure you're delivering on your KPIs, which are easy to set. I mean, that to me seems like the future, unfortunately. And it lacks the human touch, but it skyrocketed productivity. So, and at the end of the day, when it comes to capitalism, productivity is king. And that does result in better and higher incentive living for everybody. Now, that also should mean, hopefully, everyone has more time on their hands. And the products and services are less expensive, which means we can spend more time in communities and with our families and all of that stuff. pessimistic view is that we all just get wiped out and is just doing everything for us,

but there's no UBI, meaning universal basic income. There's no safety nets. And we all either have to become jugglers or some other entertainment service that can't be replaced by AI until the robots start dancing better than us. And then we're really in trouble. But, you know, I joke around with my friends that I need to go learn how to be a mechanic real quick just in case because they can't take over. Although, you know, with electric cars, even the mechanic is being kind of transformed. So it's, yeah, there's a lot. There's a lot I could say about the future.

But I think it's going to be interesting. And I think that the people who are most willing to experiment and really lead and take risks are going to be the ones on top. Yeah. But we'll see. All right. It's not about you to get a farm, you know? Yeah. Thank you.

Thank you for both the pessimistic and the optimistic. I mean, hopefully we all just have our own, you know, personal Brad Pitt, KPI therapist, you know, helping us through our daily lives. But, Ava, again, thank you for joining us on the show. Thank you if you are watching live or if you're listening to this on the podcast. As a reminder, please go to your EverydayAI.com. Sign up from the newsletter. In there, you'll learn about we're giving away two year-long premium subscriptions to check GPT. You know, so that way you can build.

your own personal Brad Pitt assistant. So thank you again for tuning in and we hope to see you back tomorrow and every day. Thank you. Thanks, Jordan. Meet Firefly AI assistant. Now live in Adobe Firefly, the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest, orchestrating multi-step workflows across Adobe Creative Cloud apps, including Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome while the assistant accelerates execution. Stand control with the ability to step in and refine at any time. See it today at firefly.adobie.com.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit Your EverydayAI.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers. and we'll see you next time.